
# Importing necessary libraries

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import re
import plotly.express as px

from sklearn.model_selection import train_test_split
from sklearn.naive_bayes import MultinomialNB
from sklearn.metrics import classification_report, confusion_matrix, ConfusionMatrixDisplay, accuracy_score, f1_score

import nltk
from nltk.corpus import stopwords
from sklearn.feature_extraction.text import CountVectorizer

nltk.download('averaged_perceptron_tagger')
from nltk import pos_tag
from nltk import RegexpParser

[nltk_data] Downloading package averaged_perceptron_tagger to
[nltk_data]     C:\Users\shrav\AppData\Roaming\nltk_data...
[nltk_data]   Package averaged_perceptron_tagger is already up-to-
[nltk_data]       date!


nltk.download('punkt')

[nltk_data] Downloading package punkt to
[nltk_data]     C:\Users\shrav\AppData\Roaming\nltk_data...
[nltk_data]   Package punkt is already up-to-date!

True


nltk.download('stopwords')

[nltk_data] Downloading package stopwords to
[nltk_data]     C:\Users\shrav\AppData\Roaming\nltk_data...
[nltk_data]   Package stopwords is already up-to-date!

True


df = pd.read_csv("spam.csv", encoding = 'ISO-8859-1')
df.head()


df.drop(df.columns[[2,3,4]], axis = 1, inplace = True)
df.columns = ['target','sms']


df.head()


df['words'] = df['sms'].apply(lambda x: len(nltk.word_tokenize(x)))
print("Word Tokens:")
print(nltk.word_tokenize(df['sms'][0]))

Word Tokens:
['Go', 'until', 'jurong', 'point', ',', 'crazy', '..', 'Available', 'only', 'in', 'bugis', 'n', 'great', 'world', 'la', 'e', 'buffet', '...', 'Cine', 'there', 'got', 'amore', 'wat', '...']


df['sentences'] = df['sms'].apply(lambda x: len(nltk.sent_tokenize(x)))
print("Sentence Tokens:")
print(nltk.sent_tokenize(df['sms'][0]))

Sentence Tokens:
['Go until jurong point, crazy..', 'Available only in bugis n great world la e buffet... Cine there got amore wat...']


df['temp'] = df['target'].map({'ham': 0, 'spam': 1})
df.head()


temp = pd.DataFrame()
temp['Target'] = ['Ham','Spam']
temp['Count'] = [len(df[df['target'] == 'ham']), len(df[df['target'] == 'spam'])]
temp = temp.sort_values(by = ['Count'], ascending = False)


fig = px.bar(temp, x = 'Target', y = 'Count', 
             color = "Target", text_auto='', width = 600,
            title = "Count Plot")

fig.update_traces(textfont_size=12, textangle = 0, textposition = "outside", cliponaxis = False)
fig.update_layout(title_x=0.5)
fig.show()


fig = px.scatter_matrix(df, dimensions=["words", "sentences"], 
                        color = "target",
                        labels={
                            "target":"Target" },
                        title = "Pair Plot" )
fig.update_layout(
    autosize=False,
    width=900,
    height=500,
    title_x=0.5
)
fig.show()


corpus = []

for i in range(len(df)):
    # Removal of punctuations, special characters
    msg = re.sub('[^a-zA-Z]', ' ', df['sms'][i]) 

    # Converting the sms to lowercase
    msg = msg.lower()
    msg = msg.split()

    # Removal of Stopwords
    msg = [word for word in msg if not word in stopwords.words('english')]

    msg = ' '.join(msg)
    corpus.append(msg)

for i in range(0, 5):
    print("Sentence-{}: {}".format(i+1,corpus[i]))

Sentence-1: go jurong point crazy available bugis n great world la e buffet cine got amore wat
Sentence-2: ok lar joking wif u oni
Sentence-3: free entry wkly comp win fa cup final tkts st may text fa receive entry question std txt rate c apply
Sentence-4: u dun say early hor u c already say
Sentence-5: nah think goes usf lives around though


df['Corpus'] = corpus
df.head()


vectorizer = CountVectorizer()
X = vectorizer.fit_transform(df['Corpus']).toarray()
Y = df['temp'].values
vectorizer.get_feature_names()

['aa',
 'aah',
 'aaniye',
 'aaooooright',
 'aathi',
 'ab',
 'abbey',
 'abdomen',
 'abeg',
 'abel',
 'aberdeen',
 'abi',
 'ability',
 'abiola',
 'abj',
 'able',
 'abnormally',
 'aboutas',
 'abroad',
 'absence',
 'absolutely',
 'absolutly',
 'abstract',
 'abt',
 'abta',
 'aburo',
 'abuse',
 'abusers',
 'ac',
 'academic',
 'acc',
 'accent',
 'accenture',
 'accept',
 'access',
 'accessible',
 'accidant',
 'accident',
 'accidentally',
 'accommodation',
 'accommodationvouchers',
 'accomodate',
 'accomodations',
 'accordin',
 'accordingly',
 'account',
 'accounting',
 'accounts',
 'accumulation',
 'achan',
 'ache',
 'achieve',
 'acid',
 'acknowledgement',
 'acl',
 'acnt',
 'aco',
 'across',
 'act',
 'acted',
 'actin',
 'acting',
 'action',
 'activ',
 'activate',
 'active',
 'activities',
 'actor',
 'actual',
 'actually',
 'ad',
 'adam',
 'add',
 'addamsfa',
 'added',
 'addicted',
 'addie',
 'adding',
 'address',
 'adds',
 'adewale',
 'adi',
 'adjustable',
 'admin',
 'administrator',
 'admirer',
 'admission',
 'admit',
 'adore',
 'adoring',
 'adp',
 'adress',
 'adrian',
 'ads',
 'adsense',
 'adult',
 'adults',
 'advance',
 'adventure',
 'adventuring',
 'advice',
 'advise',
 'advising',
 'advisors',
 'ae',
 'aeronautics',
 'aeroplane',
 'afew',
 'affair',
 'affairs',
 'affection',
 'affectionate',
 'affections',
 'affidavit',
 'afford',
 'afghanistan',
 'afraid',
 'africa',
 'african',
 'aft',
 'afternon',
 'afternoon',
 'afternoons',
 'afterwards',
 'aftr',
 'ag',
 'agalla',
 'age',
 'agency',
 'agent',
 'agents',
 'ages',
 'agidhane',
 'aging',
 'ago',
 'agree',
 'ah',
 'aha',
 'ahead',
 'ahhh',
 'ahhhh',
 'ahmad',
 'ahold',
 'aid',
 'aids',
 'aig',
 'aight',
 'aint',
 'air',
 'airport',
 'airtel',
 'aiya',
 'aiyah',
 'aiyar',
 'aiyo',
 'aj',
 'ajith',
 'ak',
 'aka',
 'akon',
 'al',
 'alaikkum',
 'alaipayuthe',
 'albi',
 'album',
 'alcohol',
 'aldrine',
 'alert',
 'alertfrom',
 'alerts',
 'aletter',
 'alex',
 'alfie',
 'algarve',
 'algebra',
 'algorithms',
 'ali',
 'alian',
 'alibi',
 'alive',
 'allah',
 'allalo',
 'allday',
 'alle',
 'allo',
 'allow',
 'allowed',
 'allows',
 'almost',
 'alone',
 'along',
 'alot',
 'already',
 'alright',
 'alrite',
 'also',
 'alter',
 'alternative',
 'although',
 'alto',
 'aluable',
 'alwa',
 'always',
 'alwys',
 'amanda',
 'amazing',
 'ambitious',
 'ambrith',
 'american',
 'ami',
 'amigos',
 'amk',
 'amla',
 'amma',
 'ammae',
 'ammo',
 'among',
 'amongst',
 'amore',
 'amount',
 'amp',
 'amplikater',
 'amrca',
 'amrita',
 'ams',
 'amt',
 'amused',
 'amy',
 'ana',
 'anal',
 'analysis',
 'anand',
 'anderson',
 'andre',
 'andres',
 'andrews',
 'andros',
 'angels',
 'angry',
 'animal',
 'animation',
 'anjie',
 'anjola',
 'anna',
 'annie',
 'anniversary',
 'annoncement',
 'announced',
 'announcement',
 'annoyin',
 'annoying',
 'anonymous',
 'anot',
 'another',
 'ans',
 'ansr',
 'answer',
 'answered',
 'answerin',
 'answering',
 'answers',
 'answr',
 'antelope',
 'antha',
 'anthony',
 'anti',
 'antibiotic',
 'anybody',
 'anyhow',
 'anymore',
 'anyone',
 'anyones',
 'anyplaces',
 'anythiing',
 'anythin',
 'anything',
 'anythingtomorrow',
 'anytime',
 'anyway',
 'anyways',
 'anywhere',
 'aom',
 'apart',
 'apartment',
 'apes',
 'apeshit',
 'aphex',
 'apnt',
 'apo',
 'apologetic',
 'apologise',
 'apologize',
 'apology',
 'app',
 'apparently',
 'appeal',
 'appear',
 'appendix',
 'applausestore',
 'apple',
 'applebees',
 'apples',
 'application',
 'apply',
 'applyed',
 'applying',
 'appointment',
 'appointments',
 'appreciate',
 'appreciated',
 'approaches',
 'approaching',
 'appropriate',
 'approve',
 'approved',
 'approx',
 'apps',
 'appt',
 'appy',
 'apr',
 'april',
 'aproach',
 'apt',
 'aptitude',
 'aq',
 'aquarius',
 'ar',
 'arab',
 'arabian',
 'arcade',
 'archive',
 'ard',
 'area',
 'arent',
 'arestaurant',
 'aretaking',
 'areyouunique',
 'argentina',
 'argh',
 'argue',
 'arguing',
 'argument',
 'arguments',
 'aries',
 'arise',
 'arises',
 'arithmetic',
 'arm',
 'armand',
 'armenia',
 'arms',
 'arng',
 'arngd',
 'arnt',
 'around',
 'aroundn',
 'arr',
 'arrange',
 'arranging',
 'arrested',
 'arrival',
 'arrive',
 'arrived',
 'arrow',
 'arsenal',
 'art',
 'artists',
 'arts',
 'arty',
 'arul',
 'arun',
 'asa',
 'asap',
 'asda',
 'ashes',
 'ashley',
 'ashwini',
 'asia',
 'asian',
 'asjesus',
 'ask',
 'askd',
 'asked',
 'askin',
 'asking',
 'asks',
 'aslamalaikkum',
 'asleep',
 'asp',
 'aspects',
 'ass',
 'assessment',
 'asshole',
 'assistance',
 'associate',
 'asssssholeeee',
 'assume',
 'assumed',
 'asthere',
 'asthma',
 'astne',
 'astoundingly',
 'astrology',
 'astronomer',
 'asus',
 'asusual',
 'ate',
 'athletic',
 'athome',
 'atlanta',
 'atlast',
 'atleast',
 'atm',
 'atrocious',
 'attach',
 'attached',
 'attack',
 'attempt',
 'atten',
 'attend',
 'attended',
 'attending',
 'attention',
 'attitude',
 'attraction',
 'attractive',
 'attracts',
 'attributed',
 'atural',
 'auction',
 'audiitions',
 'audition',
 'audrey',
 'audrie',
 'august',
 'aunt',
 'auntie',
 'aunties',
 'aunts',
 'aunty',
 'aust',
 'australia',
 'authorise',
 'auto',
 'autocorrect',
 'av',
 'ava',
 'availa',
 'available',
 'avalarr',
 'avatar',
 'avble',
 'ave',
 'avenge',
 'avent',
 'avenue',
 'avin',
 'avo',
 'avoid',
 'avoiding',
 'avoids',
 'await',
 'awaiting',
 'awake',
 'award',
 'awarded',
 'away',
 'awesome',
 'awkward',
 'aww',
 'awww',
 'ax',
 'axel',
 'axis',
 'ay',
 'ayn',
 'ayo',
 'ba',
 'baaaaaaaabe',
 'baaaaabe',
 'babe',
 'babes',
 'babies',
 'baby',
 'babygoodbye',
 'babyjontet',
 'babysit',
 'babysitting',
 'bac',
 'back',
 'backdoor',
 'backwards',
 'bad',
 'badass',
 'badly',
 'badrith',
 'bag',
 'bags',
 'bahamas',
 'baig',
 'bailiff',
 'bajarangabali',
 'bak',
 'bakra',
 'bakrid',
 'balance',
 'ball',
 'baller',
 'balloon',
 'balls',
 'bam',
 'bambling',
 'band',
 'bandages',
 'bang',
 'bangb',
 'bangbabes',
 'bani',
 'bank',
 'banks',
 'banned',
 'banneduk',
 'banter',
 'bao',
 'bar',
 'barbie',
 'barcelona',
 'bare',
 'barely',
 'bari',
 'barkleys',
 'barmed',
 'barolla',
 'barred',
 'barrel',
 'barring',
 'barry',
 'bars',
 'base',
 'based',
 'bash',
 'basic',
 'basically',
 'basket',
 'basketball',
 'basq',
 'bat',
 'batch',
 'batchlor',
 'bath',
 'bathe',
 'bathing',
 'bathroom',
 'batsman',
 'batt',
 'battery',
 'battle',
 'bawling',
 'bay',
 'bb',
 'bbc',
 'bbd',
 'bbdeluxe',
 'bbq',
 'bc',
 'bcaz',
 'bck',
 'bcm',
 'bcmsfwc',
 'bcoz',
 'bcum',
 'bcums',
 'bcz',
 'bday',
 'beach',
 'beads',
 'bear',
 'bears',
 'beatings',
 'beauties',
 'beautiful',
 'beauty',
 'bec',
 'becaus',
 'becausethey',
 'become',
 'becomes',
 'becoz',
 'becz',
 'bed',
 'bedrm',
 'bedroom',
 'beeen',
 'beehoon',
 'beendropping',
 'beer',
 'beerage',
 'beers',
 'befor',
 'beforehand',
 'beg',
 'beggar',
 'begging',
 'begin',
 'begins',
 'begun',
 'behalf',
 'behave',
 'behind',
 'bein',
 'believe',
 'belive',
 'bell',
 'bellearlier',
 'belligerent',
 'belly',
 'belong',
 'belongs',
 'belovd',
 'beloved',
 'belt',
 'ben',
 'bend',
 'beneath',
 'beneficiary',
 'benefits',
 'bennys',
 'bergkamp',
 'beside',
 'best',
 'bet',
 'beta',
 'beth',
 'betta',
 'better',
 'bettersn',
 'bettr',
 'beverage',
 'bevies',
 'beware',
 'beyond',
 'bf',
 'bffs',
 'bfore',
 'bhaji',
 'bhamb',
 'bhaskar',
 'bhayandar',
 'bian',
 'biatch',
 'bid',
 'bids',
 'big',
 'bigger',
 'biggest',
 'bike',
 'bill',
 'billed',
 'billing',
 'billion',
 'bills',
 'billy',
 'bilo',
 'bimbo',
 'bin',
 'biola',
 'bird',
 'birds',
 'birla',
 'biro',
 'birth',
 'birthdate',
 'birthday',
 'bishan',
 'bit',
 'bitch',
 'bitching',
 'bite',
 'bites',
 'bits',
 'biz',
 'bk',
 'black',
 'blackberry',
 'blacko',
 'blah',
 'blake',
 'blame',
 'blank',
 'blanked',
 'blanket',
 'blankets',
 'blastin',
 'bleak',
 'bleh',
 'bless',
 'blessed',
 'blessing',
 'blessings',
 'blimey',
 'blind',
 'block',
 'blocked',
 'blog',
 'blogging',
 'blogspot',
 'bloke',
 'blokes',
 'blonde',
 'bloo',
 'blood',
 'bloody',
 'bloomberg',
 'blow',
 'blowing',
 'blown',
 'blu',
 'blue',
 'bluetooth',
 'bluetoothhdset',
 'bluff',
 'blur',
 'bluray',
 'bmw',
 'board',
 'boat',
 'boatin',
 'bob',
 'body',
 'boggy',
 'bognor',
 'bold',
 'bollox',
 'boltblue',
 'bomb',
 'bone',
 'bong',
 'bonus',
 'boo',
 'boobs',
 'book',
 'booked',
 'bookedthe',
 'booking',
 'bookmark',
 'books',
 'bookshelf',
 'boooo',
 'boost',
 'booty',
 'bootydelious',
 'borderline',
 'bored',
 'borin',
 'boring',
 'born',
 'borrow',
 'boss',
 'boston',
 'bot',
 'bother',
 'bothering',
 'bottle',
 'bottom',
 'bought',
 'boundaries',
 'bout',
 'bowa',
 'bowl',
 'bowls',
 'box',
 'boy',
 'boye',
 'boyf',
 'boyfriend',
 'boys',
 'boytoy',
 'bp',
 'bpo',
 'brah',
 'brain',
 'braindance',
 'brainless',
 'brains',
 'brainy',
 'brand',
 'brandy',
 'bras',
 'brats',
 'braved',
 'bray',
 'brb',
 'brdget',
 'bread',
 'breadstick',
 'break',
 'breaker',
 'breakfast',
 'breakin',
 'breaking',
 'breaks',
 'breath',
 'breathe',
 'breather',
 'breathing',
 'breeze',
 'breezy',
 'bremoved',
 'bribe',
 'bridal',
 'bridge',
 'bridgwater',
 'brief',
 'bright',
 'brighten',
 'brilliant',
 'brilliantly',
 'brin',
 'bring',
 'bringing',
 'brings',
 'brisk',
 'brison',
 'bristol',
 'british',
 'britney',
 'bro',
 'broad',
 'broadband',
 'broke',
 'broken',
 'brolly',
 'bros',
 'broth',
 'brothas',
 'brother',
 'brothers',
 'brought',
 'brown',
 'brownie',
 'brownies',
 'browse',
 'browser',
 'browsin',
 'bruce',
 'brum',
 'bruv',
 'bslvyl',
 'bsn',
 'bsnl',
 'bstfrnd',
 'bt',
 'btw',
 'btwn',
 'bubbletext',
 'bucks',
 'bud',
 'buddy',
 'buddys',
 'budget',
 'buen',
 'buff',
 'buffet',
 'buffy',
 'bugis',
 'build',
 'building',
 'built',
 'bulbs',
 'bull',
 'bullshit',
 'bunch',
 'bundle',
 'bunkers',
 'buns',
 'burden',
 'burger',
 'burgundy',
 'burial',
 'burn',
 'burning',
 'burns',
 'burnt',
 'burrito',
 'bus',
 'buses',
 'busetop',
 'business',
 'busty',
 'busy',
 'butt',
 'buttheres',
 'butting',
 'buttons',
 'buy',
 'buyer',
 'buyers',
 'buying',
 'buzy',
 'buzz',
 'buzzzz',
 'bw',
 'bx',
 'byatch',
 'bye',
 'cab',
 'cabin',
 'cable',
 'cafe',
 'cage',
 'cake',
 'cakes',
 'cal',
 'calculated',
 'calculation',
 'cali',
 'calicut',
 'california',
 'call',
 'callback',
 'callcost',
 'calld',
 'called',
 'caller',
 'callers',
 'callertune',
 'callfreefone',
 'callin',
 'calling',
 'callon',
 'calls',
 'calm',
 'cam',
 'camcorder',
 'came',
 'camera',
 'camp',
 'campus',
 'camry',
 'canada',
 'canal',
 'canary',
 'cancel',
 'canceled',
 'cancelled',
 'cancer',
 'canlove',
 'cann',
 'canname',
 'cannot',
 'cant',
 'cantdo',
 'canteen',
 'capacity',
 'capital',
 'cappuccino',
 'caps',
 'captain',
 'captaining',
 'car',
 'card',
 'cardiff',
 'cardin',
 'cards',
 'care',
 'careabout',
 'cared',
 'career',
 'careers',
 'careful',
 'carefully',
 'careless',
 'cares',
 'caring',
 'carlie',
 'carlin',
 'carlos',
 'carly',
 'carolina',
 'caroline',
 'carpark',
 'carry',
 'carryin',
 'cars',
 'cartons',
 'cartoon',
 'case',
 'cash',
 'cashbin',
 'cashed',
 'cashto',
 'casing',
 'cast',
 'casting',
 'castor',
 'casualty',
 'cat',
 'catch',
 'catches',
 'catching',
 ...]


x_train, x_test, y_train, y_test = train_test_split(X, Y, test_size = 0.2, random_state = 42)

print("Size of X,y in train and test data")
print(f"X train: {x_train.shape} \nX test:  {x_test.shape} \ny train: {y_train.shape} \ny test:  {y_test.shape}")

Size of X,y in train and test data
X train: (4457, 7546) 
X test:  (1115, 7546) 
y train: (4457,) 
y test:  (1115,)


nb = MultinomialNB()
nb.fit(x_train, y_train)
yp_train = nb.predict(x_train)
yp_test = nb.predict(x_test)


cm_train = confusion_matrix(y_train, yp_train)
train_cmd = ConfusionMatrixDisplay(cm_train)
s1 = round((cm_train[0,0]/(cm_train[0,0] + cm_train[0,1])),4)

print("**************************************************************************")
print("Classification Report for TRAIN DATA\n")
print(classification_report(y_train, yp_train))
print("Specificity on Train Data: ", s1)
print("Accuracy on Train Data: ", round(accuracy_score(y_train, yp_train),4))
print("F1 Score on Train Data: ", round(f1_score(y_train, yp_train),4))
print("**************************************************************************")

**************************************************************************
Classification Report for TRAIN DATA

              precision    recall  f1-score   support

           0       1.00      1.00      1.00      3860
           1       0.97      0.98      0.97       597

    accuracy                           0.99      4457
   macro avg       0.98      0.99      0.99      4457
weighted avg       0.99      0.99      0.99      4457

Specificity on Train Data:  0.9953
Accuracy on Train Data:  0.993
F1 Score on Train Data:  0.9741

**************************************************************************


cm_test = confusion_matrix(y_test, yp_test)
test_cmd = ConfusionMatrixDisplay(cm_test)
s2 = round((cm_test[0,0]/(cm_test[0,0] + cm_test[0,1])),4)

print("**************************************************************************")
print("\nClassification Report for TEST DATA\n")
print(classification_report(y_test, yp_test))
print("Specificity on Test Data: ", s2)
print("Accuracy on Test Data: ", round(accuracy_score(y_test, yp_test), 4))
print("F1 Score Test Data: ", round(f1_score(y_test, yp_test), 4))
print("**************************************************************************")

**************************************************************************

Classification Report for TEST DATA

              precision    recall  f1-score   support

           0       0.99      0.98      0.98       965
           1       0.86      0.94      0.90       150

    accuracy                           0.97      1115
   macro avg       0.93      0.96      0.94      1115
weighted avg       0.97      0.97      0.97      1115

Specificity on Test Data:  0.9762
Accuracy on Test Data:  0.9713
F1 Score Test Data:  0.8981
**************************************************************************


print("Confusion Matrix for Train Data")
train_cmd.plot()

Confusion Matrix for Train Data

<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x1b8cf39db80>


print("Confusion Matrix for Test Data")
test_cmd.plot()

Confusion Matrix for Test Data

<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x1b8cf57cfa0>


df.head()


first_four_rows=list(df['Corpus'][:4])


print("POS tagging on first 4 rows of 'sms'")
for i,sent in enumerate(first_four_rows):
    tokens_tag = pos_tag(sent.split())
    print(f"Sentence {i+1} : {tokens_tag} ")

POS tagging on first 4 rows of 'sms'
Sentence 1 : [('go', 'VB'), ('jurong', 'JJ'), ('point', 'NN'), ('crazy', 'NN'), ('available', 'JJ'), ('bugis', 'NN'), ('n', 'RB'), ('great', 'JJ'), ('world', 'NN'), ('la', 'NN'), ('e', 'VBP'), ('buffet', 'JJ'), ('cine', 'NN'), ('got', 'VBD'), ('amore', 'RB'), ('wat', 'JJ')] 
Sentence 2 : [('ok', 'JJ'), ('lar', 'JJ'), ('joking', 'NN'), ('wif', 'NN'), ('u', 'JJ'), ('oni', 'NN')] 
Sentence 3 : [('free', 'JJ'), ('entry', 'NN'), ('wkly', 'VBD'), ('comp', 'NN'), ('win', 'NN'), ('fa', 'JJ'), ('cup', 'VBZ'), ('final', 'JJ'), ('tkts', 'NN'), ('st', 'NN'), ('may', 'MD'), ('text', 'VB'), ('fa', 'JJ'), ('receive', 'JJ'), ('entry', 'NN'), ('question', 'NN'), ('std', 'VBD'), ('txt', 'JJ'), ('rate', 'NN'), ('c', 'NNS'), ('apply', 'VBP')] 
Sentence 4 : [('u', 'JJ'), ('dun', 'NNS'), ('say', 'VBP'), ('early', 'JJ'), ('hor', 'NN'), ('u', 'JJ'), ('c', 'NN'), ('already', 'RB'), ('say', 'VB')]


import spacy
nlp=spacy.load('en_core_web_sm')
from spacy import displacy
txt=nlp(u"the series opened 17 years later, as Viserys Targaryen tried to win an eastern tribal army to his side, so he could retake the Iron Throne")

displacy.render(txt,jupyter=True)


token_data=[]
for token in txt:
  token_row_data = [token.text, token.pos_, token.dep_, (token.head.text, token.head.i), token.head.pos_, 
                    [child.i for child in token.children]]
  token_data.append(token_row_data)

token_data_df = pd.DataFrame(token_data, columns=['text', 'POS', 'dependency', 'head', 'head_POS', 'children'])
token_data_df.head()

	v1	v2	Unnamed: 2	Unnamed: 3	Unnamed: 4
0	ham	Go until jurong point, crazy.. Available only ...	NaN	NaN	NaN
1	ham	Ok lar... Joking wif u oni...	NaN	NaN	NaN
2	spam	Free entry in 2 a wkly comp to win FA Cup fina...	NaN	NaN	NaN
3	ham	U dun say so early hor... U c already then say...	NaN	NaN	NaN
4	ham	Nah I don't think he goes to usf, he lives aro...	NaN	NaN	NaN

	text	POS	dependency	head	head_POS	children
0	the	DET	det	(series, 1)	NOUN	[]
1	series	NOUN	nsubj	(opened, 2)	VERB	[0]
2	opened	VERB	ccomp	(retake, 24)	VERB	[1, 5, 6, 10]
3	17	NUM	nummod	(years, 4)	NOUN	[]
4	years	NOUN	npadvmod	(later, 5)	ADV	[3]

Exploratory Data Analysis (EDA)¶

Creating Corpus¶

Model building¶

Multinomial NB model¶

	target	sms	words	sentences	temp	Corpus
0	ham	Go until jurong point, crazy.. Available only ...	24	2	0	go jurong point crazy available bugis n great ...
1	ham	Ok lar... Joking wif u oni...	8	2	0	ok lar joking wif u oni
2	spam	Free entry in 2 a wkly comp to win FA Cup fina...	37	2	1	free entry wkly comp win fa cup final tkts st ...
3	ham	U dun say so early hor... U c already then say...	13	1	0	u dun say early hor u c already say
4	ham	Nah I don't think he goes to usf, he lives aro...	15	1	0	nah think goes usf lives around though